Reply to this post | Go Back

View Post [edit]

Poster:	me#1	Date:	March 04, 2006 12:12:06am
Forum:	web	Subject:	Archived XHTML broken

Archived XHTML is broken because of the inserted unclosed *BASE HREF="..."*
It should be *base href="..."/* (lowercase and with the slash).
See http://web.archive.org/www.w3.org/People/mimasa/test/xhtml/media-types/ for an example.

I've also seen JavaScript inserted using *SCRIPT language="Javascript"* when the *base href="..."/* was already provided in the original document. That should be *script type="application/javascript"* (MIME type is registered in RFC 4329.) But a (XML compliant) rewriting of the *base/* element would better, I think.

[For visibility, all > and < enclosures above have been replaced with *. - mod]

This post was modified by Diana Hamilton on 2006-03-04 08:12:06

Reply to this post

Reply [edit]

Poster:	gojomo	Date:	March 06, 2006 08:32:23am
Forum:	web	Subject:	Re: Archived XHTML broken

I've added a bug report to ensure that a future version of the Wayback Machine corrects these issues.

In the meantime, could you help me determine the severity of the issue? Do these issues cause unrecoverable problems (visual or functional) for browsers, viewers, or other client software expecting valid XHTML? If so, can you provide an example of a page or pages from the archive so affected?

Thanks for your report,

- Gordon @ IA

Reply to this post

Reply [edit]

Poster:	me#1	Date:	March 07, 2006 06:17:24am
Forum:	web	Subject:	Re: Archived XHTML broken

Hi,
just view http://web.archive.org/www.w3.org/People/mimasa/test/xhtml/media-types/ with an XHTML compatible browser like Firefox or Opera. They only show an XML parse error because it's not "well-formed". Internet Explorer prompts to download the page, but that's because it doesn't support application/xhtml+xml, that also occurs on live pages.

The page with the inserted JavaScript (see http://web.archive.org/http://geckoblaster.de/ for an example) is displayed in the browser (in Firefox and Opera at least), but Firefox does not execute the script because
1. *SCRIPT/* is an unknown tag in XHTML, there's only *script/*
2. it's commented out with *!-- --*.
Opera executes the script, but I'd say that Opera shouldn't.
However, the script wouldn't be necessary if just the *base href="..."/* was rewritten.

And by the way, stylesheets included by an XML procession instruction like *?xml-stylesheet href="style.css" type="text/css"?* are not archived.

[For visibility, all > and < enclosures above have been replaced with *. - mod]

This post was modified by Diana Hamilton on 2006-03-07 14:17:24